source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/GaussianProcess/CovarianceProd.cs @ 8612

Last change on this file since 8612 was 8612, checked in by gkronber, 7 years ago

#1902 implemented all mean and covariance functions with parameters as ParameterizedNamedItems

File size: 3.5 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2012 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
4 *
5 * This file is part of HeuristicLab.
6 *
7 * HeuristicLab is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation, either version 3 of the License, or
10 * (at your option) any later version.
11 *
12 * HeuristicLab is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15 * GNU General Public License for more details.
16 *
17 * You should have received a copy of the GNU General Public License
18 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
19 */
20#endregion
21
22using System;
23using System.Collections.Generic;
24using System.Linq;
25using HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
28
29namespace HeuristicLab.Algorithms.DataAnalysis {
30  [StorableClass]
31  [Item(Name = "CovarianceProd",
32    Description = "Product covariance function for Gaussian processes.")]
33  public sealed class CovarianceProd : Item, ICovarianceFunction {
34    [Storable]
35    private ItemList<ICovarianceFunction> factors;
36
37    [Storable]
38    private int numberOfVariables;
39    public ItemList<ICovarianceFunction> Factors {
40      get { return factors; }
41    }
42
43    [StorableConstructor]
44    private CovarianceProd(bool deserializing)
45      : base(deserializing) {
46    }
47
48    private CovarianceProd(CovarianceProd original, Cloner cloner)
49      : base(original, cloner) {
50      this.factors = cloner.Clone(original.factors);
51      this.numberOfVariables = original.numberOfVariables;
52    }
53
54    public CovarianceProd()
55      : base() {
56      this.factors = new ItemList<ICovarianceFunction>();
57    }
58
59    public override IDeepCloneable Clone(Cloner cloner) {
60      return new CovarianceProd(this, cloner);
61    }
62
63    public int GetNumberOfParameters(int numberOfVariables) {
64      this.numberOfVariables = numberOfVariables;
65      return factors.Select(f => f.GetNumberOfParameters(numberOfVariables)).Sum();
66    }
67
68    public void SetParameter(double[] hyp) {
69      if (factors.Count == 0) throw new ArgumentException("at least one factor is necessary for the product covariance function.");
70      int offset = 0;
71      foreach (var t in factors) {
72        var numberOfParameters = t.GetNumberOfParameters(numberOfVariables);
73        t.SetParameter(hyp.Skip(offset).Take(numberOfParameters).ToArray());
74        offset += numberOfParameters;
75      }
76    }
77
78    public double GetCovariance(double[,] x, int i, int j) {
79      return factors.Select(f => f.GetCovariance(x, i, j)).Aggregate((a, b) => a * b);
80    }
81
82    public IEnumerable<double> GetGradient(double[,] x, int i, int j) {
83      var covariances = factors.Select(f => f.GetCovariance(x, i, j)).ToArray();
84      for (int ii = 0; ii < factors.Count; ii++) {
85        foreach (var g in factors[ii].GetGradient(x, i, j)) {
86          double res = g;
87          for (int jj = 0; jj < covariances.Length; jj++)
88            if (ii != jj) res *= covariances[jj];
89          yield return res;
90        }
91      }
92    }
93
94    public double GetCrossCovariance(double[,] x, double[,] xt, int i, int j) {
95      return factors.Select(f => f.GetCrossCovariance(x, xt, i, j)).Aggregate((a, b) => a * b);
96    }
97  }
98}
Note: See TracBrowser for help on using the repository browser.